The major purpose of this study is to investigate the generalizability of results from the experimental laboratory to the general population and its component sub-groups. To that end, a number of systematic replications of experiments were conducted with respondents in their homes during a household sample survey. The experiments, in the substantive areas of social reinforcement reward distribution, and attribution processes, satisfied or approached the following four criteria; (1) experimental studies have produced consistent and replicable results; (2) restricted sampling bases have been utilized; (3) experimental results have been relied upon to explain or interpret behavior of individuals with differing social characteristics than those of the sample producing the experimental data; and (4) data indicate that these social characteristics might interact with experimental situations, such that generalization becomes tenuous. The nature of the experimental procedures permits the examination of issues related to generalizability beyond those of population differences. In addition to the inclusion of manipulation and measures identical to those utilized in prior laboratory studies, novel manipulations and the random assignment of conditions and interviewers allow the examination of new hypotheses of substantive interest within each of the experiments. The proposal requests support to analyze, interpret, and write-up data gathered in the 1972-73 University of Michigan Detroit Area Study. The multi-stage random sample of the Detroit Metropolitan area consists of 576 respondents. The procedures, randomization of conditions and interviewers, multiple experiments, sample selection and size, provide high quality data uniquely suited to the objectives of the study. Thorough analysis of the data should provide information regarding a number of important issues. Among these are: (1) an empirical assessment of the generalizability and robustness of a sample of past laboratory research; (2) applicability and suitability of the method for investigating new hypotheses; and (3) an empirical basis for examining the feasibility of combining two respected methods (surveys and laboratory research) of data gathering in the behavioral sciences.